RECENT FINDINGS:[unreadable] THE FUNCTION AND STRUCTURE OF ORF1p - ORF1p is a multimer, the monomer of which has three domains: A rapidly evolving amino terminal domain of unknown function; a coiled coil domain, which is required for multimer formation and which also evolves rapidly having undergone several episodes of adaptive evolution; and a highly conserved carboxy-terminal domain that binds nucleic acids. These structural and evolutionary characteristics are typical of all the mammalian L1 ORF1p's so far examined. Thus, rapid evolution of the amino terminal half of ORF1p must be essential for the persistence of L1 activity during the evolution of mammals. To understand why, we are determining the effects of ORF1p evolution on its structure, on retrotransposition, and on a number of ORF1p properties shown to be essential for retrotransposition: e.g., RNA-binding, multimer formation, and nucleic acid chaperone activity. (Also see Z01 DK057811-01 LMCB: Mammalian L1 Retrotransposon - host interaction). We constructed an ancestral version of ORF1p (L1Pa5) that predated the evolutionary changes of the modern version (L1Pa1) and mosaic ORF1p's that contain modern and ancestral regions. We implemented baculovirus production of ORF1p after having shown that expression in E. coli was unsuitable due to improper translation of the ORF1p transcript. Baculovirus expression provides ample protein for both biochemical and structural analysis. Both the multicoil structural prediction program and analytical centrifugation showed that unlike mouse ORF1p, the modern human ORF1p is either a dimer or a very labile trimer. In contrast, multicoil predicts that the pre-adapted ancestral coiled coil will form a strong trimer. Multicoil predictions of all available mammalian non-human ORF1p amino acid sequences showed that strong trimer formation is the rule. Substitution of just a few ancestral amino acids for their modern counterparts in a small region of the coiled coil domain completely abolishes retrotransposition and converts the multicoil prediction from dimer / weak trimer to strong trimer. But restoration of just a subset of these ancestral amino acids to their modern counterparts restores retrotransposition and the prediction of dimer / weak trimer. Phylogenetic analysis of both the evolutionary antecedents of ORF1p in the human lineage and of ORF1p from different species showed that, aside from those amino acids that underwent adaptive change (positive selection), this region of the coiled coil domain has otherwise exhibited very little evolutionary change. We are now implementing baculovirus expression of our various ORF1p proteins for biochemical and structural analysis. We are also continuing our collaboration with Fred Dyda and Allison Hickman, structural biologists in NIDDK to determine their structures. We are now carrying out nucleic acid binding experiments preliminary to attempting to co-crystallizing ORF1p with RNA or DNA, as we could not obtain crystals with just the protein.[unreadable] DEVELOPMENT OF A NEW RETROTRANSPOSITION ASSAY - Retrotransposition assays generally rely on the detection of a DNA copy (reporter gene) of a spliced RNA transcript. The requirement for RNA splicing ensures that DNA copy has gone through an RNA intermediate, a hallmark of retrotransposition. There are three major problems with the current retrotransposition assay: First - The splicing reaction occurs via the normal splicesomal pathway, a pathway that L1 transcripts do not normally encounter as L1 RNA is not spliced. Spliced RNAs bind proteins unique to this pathway and have dramatically different metabolic fates from non-spliced RNAs. We rectified this problem by substituting a self-splicing intron for the splicesomal intron. Although the two vectors generate nearly identical inserts, they differ in their retrotransposition activity, a difference that seemingly is greatly enhanced when the L1 element contains the ancestral ORF1p. Second - The integrity of the L1 3' UTR RNA is severely compromised by the presence of the 2 kb reporter gene. Thus, any role of the highly conserved sequences and sequence motifs present in the 200 bp 3' UTR RNA could be overwhelmed by its juxtaposition to 2000 bp of extraneous RNA. We hope to address this problem by replacing the reporter gene with a 15 bp sequence split by the self-splicing (autocatalytic) intron. Thus, as soon as the transcript is synthesized, the intron is spliced out generating an L1 transcript that differs from a normal L1 transcript by only 15 nucleotides. And by inserting the reporter in the least conserved region of the 3' UTR we hope to minimize its effect on L1 RNA. Third, the present assay does not reveal information on the intermediate steps in L1 retrotransposition; i.e., L1 transcript and cDNA levels. To address this issue we will isolate the various retrotransposition intermediates and products by hybridization with the appropriate complementary PNA sequence (a nucleic acid polymer in which the phoshpodiester backbone is replace by amide bonds) and then quantify the products by real time (RT) PCR. We are collaborating in this effort with Dr. Daniel Appella (NIDDK) who has developed modifications to PNA synthesis that greatly increase both its sensitivity and specificity.